Getting Started
Resources
Exercises taken from:
https://github.com/jennybc/r-graph-catalog
which are based on:
Creating more effective graphs
Exercise 1 - Rank the categories
How did you do?
Exercise 2 - Distance - common base
How did you do?
Exercise 3 - Area and Volume
How did you do?
Exercise 4 - Length
How did you do?
That's why stacked bar charts are never recommended
Finally colour, hue, saturation, density
Finally colour, hue, saturation, density (cont…)
Finally colour, hue, saturation, density (cont…)
Finally colour, hue, saturation, density (cont…)
## End of principles session…
praise()
## [1] "You are magnificent!"
Can we apply Cleveland's and Robbin's principles to multidimentional data?
Make the data stand out. Avoid superfluity
Use visually prominent graphical elements to show the data
Do not clutter the data region
Do not overdo the number of tick marks
Use a reference line when there is an important value that must be seen across the entire graph, but do not let that line interfere with the data
Overlapping plotting symbols must be visually distinguishable
Superimposed datasets must be readily visually discriminated
Visual clarity must be preserved under reduction and reproduction
Accessing the NIWA climate database
public_user <- cf_user() # Defaults to "public" #select using menus #out <- cf_datatype()
or programatically
wind_dt <- cf_datatype(2, 1, 4, 1) # Daily Rain rain_dt <- cf_datatype(3, 1, 1) # Daily temperature extremes temp_dt <- cf_datatype(4, 2, 1) all_dts <- wind_dt + rain_dt + temp_dt all_dts
## dt.name dt.type dt.options dt.combo ## dt1 Wind Surface wind [9amWind] m/s ## dt2 Precipitation Rain (fixed periods) [Daily ] ## dt3 Temperature and Humidity Max_min_temp [DailyMaxMin]
reefton_st = cf_station() reefton_st
## name network agent start end open distance lat ## 1) Reefton Ews F21182 3925 1960-08-01 2017-05-25 TRUE 0 -42.11578 ## lon ## 1) 171.8601
Reefton Data
reefton_data <- cf_query(public_user, all_dts, reefton_st, paste(as.Date(Sys.time()) - 333, "9"))
## connecting to CliFlo...
## reading data...
## UserName is = public ## Number of charged rows output = 0 ## Number of free rows output = 1000 ## Total number of rows output = 1000 ## Note: The end date was revised to meet the maximum number of rows allowed per query [1000] ## or due to running out of rows in your subscription. Also, one or more datatypes may have been disabled due to the above. ## Copyright NIWA 2017 Subject to NIWA's Terms and Conditions ## See: http://cliflo.niwa.co.nz/pls/niwp/doc/terms.html ## Comments to: cliflo@niwa.co.nz
str(reefton_data)
## Formal class 'cfDataList' [package "clifro"] with 1 slot ## ..@ .Data:List of 3 ## .. ..$ :Formal class 'cfWind' [package "clifro"] with 6 slots ## .. .. .. ..@ .Data :List of 8 ## .. .. .. .. ..$ : Factor w/ 1 level "Reefton Ews": 1 1 1 1 1 1 1 1 1 1 ... ## .. .. .. .. ..$ : Factor w/ 334 levels "20160625:0900",..: 1 2 3 4 5 6 7 8 9 10 ... ## .. .. .. .. ..$ : int [1:334] 109 193 7 117 109 180 315 283 254 129 ... ## .. .. .. .. ..$ : num [1:334] 2.6 0.8 0.4 1.2 2.1 1 0.4 0.4 0.4 0.6 ... ## .. .. .. .. ..$ : num [1:334] 17 85 75 17 44 52 0 45 57 59 ... ## .. .. .. .. ..$ : num [1:334] 0.7 0.4 0.3 0.5 1 0.5 0.3 0.4 0.3 0.4 ... ## .. .. .. .. ..$ : int [1:334] 1 1 1 1 1 1 1 1 1 1 ... ## .. .. .. .. ..$ : Factor w/ 1 level "H": 1 1 1 1 1 1 1 1 1 1 ... ## .. .. .. ..@ data_label: chr " 9am only surface wind (m/s)" ## .. .. .. ..@ dt_name : chr "Surface Wind" ## .. .. .. ..@ dt_type : chr " 9am only" ## .. .. .. ..@ names : chr [1:8] "Station" "Date(local)" "Dir(DegT)" "Speed(m/s)" ... ## .. .. .. ..@ row.names : chr [1:334] "1" "2" "3" "4" ... ## .. ..$ :Formal class 'cfRain' [package "clifro"] with 6 slots ## .. .. .. ..@ .Data :List of 8 ## .. .. .. .. ..$ : Factor w/ 1 level "Reefton Ews": 1 1 1 1 1 1 1 1 1 1 ... ## .. .. .. .. ..$ : Factor w/ 334 levels "20160625:0900",..: 1 2 3 4 5 6 7 8 9 10 ... ## .. .. .. .. ..$ : num [1:334] 1.6 5.6 12.2 0.2 0.4 0 0 0.2 0.2 0 ... ## .. .. .. .. ..$ : logi [1:334] NA NA NA NA NA NA ... ## .. .. .. .. ..$ : num [1:334] 0 0 0 0 0 0.2 0.4 0.6 0.7 1 ... ## .. .. .. .. ..$ : num [1:334] 1.4 5.4 12 0 0.1 0 0 0 0 0 ... ## .. .. .. .. ..$ : int [1:334] 24 24 24 24 24 24 24 24 24 24 ... ## .. .. .. .. ..$ : Factor w/ 1 level "D": 1 1 1 1 1 1 1 1 1 1 ... ## .. .. .. ..@ data_label: chr " Daily Rain (mm)" ## .. .. .. ..@ dt_name : chr "Rain" ## .. .. .. ..@ dt_type : chr " Daily" ## .. .. .. ..@ names : chr [1:8] "Station" "Date(local)" "Amount(mm)" "SofG" ... ## .. .. .. ..@ row.names : chr [1:334] "1" "2" "3" "4" ... ## .. ..$ :Formal class 'cfTemp' [package "clifro"] with 7 slots ## .. .. .. ..@ .Data :List of 12 ## .. .. .. .. ..$ : Factor w/ 1 level "Reefton Ews": 1 1 1 1 1 1 1 1 1 1 ... ## .. .. .. .. ..$ : Factor w/ 332 levels "20160625:0900",..: 1 2 3 4 5 6 7 8 9 10 ... ## .. .. .. .. ..$ : num [1:332] 13.1 11.8 14.2 14.5 13.8 10.5 14.5 9.8 11.6 9.6 ... ## .. .. .. .. ..$ : int [1:332] 24 24 24 24 24 24 24 24 24 24 ... ## .. .. .. .. ..$ : num [1:332] 6.2 5.9 6.2 7.7 5.3 1.5 -3.2 -2.9 1.4 2.1 ... ## .. .. .. .. ..$ : int [1:332] 24 24 24 24 24 24 24 24 24 24 ... ## .. .. .. .. ..$ : num [1:332] 4 3.1 5.6 3 1.6 -0.8 -6.6 -2.7 -0.8 2.8 ... ## .. .. .. .. ..$ : int [1:332] 24 24 24 24 24 24 24 24 24 24 ... ## .. .. .. .. ..$ : logi [1:332] NA NA NA NA NA NA ... ## .. .. .. .. ..$ : logi [1:332] NA NA NA NA NA NA ... ## .. .. .. .. ..$ : logi [1:332] NA NA NA NA NA NA ... ## .. .. .. .. ..$ : Factor w/ 1 level "D": 1 1 1 1 1 1 1 1 1 1 ... ## .. .. .. ..@ data_label: chr " Daily maximum/minimum temperature" ## .. .. .. ..@ plot_label: language " Daily" ~ temperature ~ (degree * C) ## .. .. .. ..@ dt_name : chr "Max_min" ## .. .. .. ..@ dt_type : chr " Daily" ## .. .. .. ..@ names : chr [1:12] "Station" "Date(local)" "Tmax(C)" "Period(Hrs)" ... ## .. .. .. ..@ row.names : chr [1:332] "1" "2" "3" "4" ...
From list to a dataframe
reefton_data_df <- reefton_data %>% map_df(`[`)
## Warning in bind_rows_(x, .id): Unequal factor levels: coercing to character
head(reefton_data_df)
## Station Date.local Dir.DegT Speed.ms Dir StdDev Spd StdDev ## 1 Reefton Ews 2016-06-25 09:00:00 109 2.6 17 0.7 ## 2 Reefton Ews 2016-06-26 09:00:00 193 0.8 85 0.4 ## 3 Reefton Ews 2016-06-27 09:00:00 7 0.4 75 0.3 ## 4 Reefton Ews 2016-06-28 09:00:00 117 1.2 17 0.5 ## 5 Reefton Ews 2016-06-29 09:00:00 109 2.1 44 1.0 ## 6 Reefton Ews 2016-06-30 09:00:00 180 1.0 52 0.5 ## Period.Hrs Freq Amount.mm SofG Deficit.mm Runoff.mm Tmax.C Tmin.C ## 1 1 H NA NA NA NA NA NA ## 2 1 H NA NA NA NA NA NA ## 3 1 H NA NA NA NA NA NA ## 4 1 H NA NA NA NA NA NA ## 5 1 H NA NA NA NA NA NA ## 6 1 H NA NA NA NA NA NA ## Period.Hrs.1 Tgmin.C Period.Hrs.2 Tmean.C RHmean.% Period.Hrs.3 ## 1 NA NA NA NA NA NA ## 2 NA NA NA NA NA NA ## 3 NA NA NA NA NA NA ## 4 NA NA NA NA NA NA ## 5 NA NA NA NA NA NA ## 6 NA NA NA NA NA NA
From wide to long for plotting
reefton_data_long <- reefton_data_df %>% select(Station,Date.local,Amount.mm, Tmin.C,Speed.ms) %>% gather(measure,data,-Station,-Date.local) %>% filter(!is.na(data)) reefton_data_long$Year <- format(reefton_data_long$Date.local, "%Y")
Let's plot
(p_initial <- ggplot(reefton_data_long,aes(Date.local,data))+ geom_point(aes(colour=measure)))
Let's apply Cleveland's principles…
What things are superflous?
Let's apply Cleveland's principles…
Do the graphical elements clearly show the data?
Let's apply Cleveland's principles…
Are the number of tick marks ok?
Let's apply Cleveland's principles…
Are overlapping plotting symbols visually distinguisable?
Let's apply Cleveland's principles…
Are superimposed datasets visually distinguisable?
How should we fix this graph?
Separate superimposed dataset
ggplot(reefton_data_long,aes(Date.local,data))+ geom_point(aes(colour=measure)) + facet_grid(measure~.) #separate datasets
Common y scale not needed
ggplot(reefton_data_long,aes(Date.local,data))+ geom_point(aes(colour=measure)) + facet_grid(measure~., scales = "free_y") #variable y-scale
Make overlapping data easier to distinguish
ggplot(reefton_data_long,aes(Date.local,data))+ geom_point(aes(colour=measure),shape=1) + #show data overlap facet_grid(measure~., scales = "free")
Remove superfluous legend
ggplot(reefton_data_long,aes(Date.local,data))+ geom_point(aes(colour=measure),shape=1) + facet_grid(measure~., scales = "free_y") + theme(legend.position="none") #legend not needed
Move labels to outside the y axis
ggplot(reefton_data_long,aes(Date.local,data)) + geom_point(aes(colour=measure),shape=1) + facet_grid(measure~., scales = "free_y", switch="y") + #label on left size theme(legend.position="none",strip.placement = "outside") #label outside of y axis
Label x and y axis meaningfully
ggplot(reefton_data_long,aes(Date.local,data)) +
geom_point(aes(colour=measure),shape=1) +
facet_grid(measure~., scales = "free_y", switch="y") +
theme(legend.position="none",strip.placement = "outside") +
ylab("") + #label superfluous
xlab("Date") #make label more meaningful
Remove extra shading
ggplot(reefton_data_long,aes(Date.local,data)) +
geom_point(aes(colour=measure),shape=1) +
facet_grid(measure~., scales = "free_y", switch="y") +
theme(legend.position="none",strip.placement = "outside") +
ylab("") +
xlab("Date") +
theme_bw() #overrides previous theme
Remove extra shading
ggplot(reefton_data_long,aes(Date.local,data)) +
geom_point(aes(colour=measure),shape=1) +
facet_grid(measure~., scales = "free_y", switch="y") +
ylab("") +
xlab("Date") +
theme_bw() +
theme(legend.position="none",strip.placement = "outside") #legend not needed; move strip to outside of y axis
Change format of dates
p1 <- (
ggplot(reefton_data_long,aes(as.Date(Date.local),data)) + #explicit date class
geom_point(aes(colour=measure),shape=1) +
facet_grid(measure~., scales = "free_y", switch="y") +
ylab("") +
xlab("Date") +
theme_bw() +
theme(legend.position="none",strip.placement = "outside")+
scale_x_date(date_labels = "%m-%Y") #format of date
)
Change format of dates
Change format of dates
p1 <- (
ggplot(reefton_data_long,aes(as.Date(Date.local),data)) +
geom_point(aes(colour=measure),shape=1) +
facet_grid(measure~., scales = "free_y", switch="y") +
ylab("") +
theme_bw() +
theme(legend.position="none",strip.placement = "outside")+
scale_x_date(date_labels = "%B %Y") #format of date
)
Change format of dates
Changing the labels
reefton_data_long$measure <- factor(reefton_data_long$measure,
labels = c("Rainfall~amount~(mm)",
"Wind~speed~(ms^-1)",
"Min~daily~temp~(degree*C)"
))
p1 <- (
ggplot(reefton_data_long,aes(as.Date(Date.local),data)) +
geom_point(aes(colour=measure),shape=1) +
facet_grid(measure~., scales = "free_y", switch="y",
labeller = label_parsed) + # parses math expressions
ylab("") +
xlab("Date") +
theme_bw() +
theme(legend.position="none",strip.placement = "outside") +
scale_x_date(date_labels = "%B %Y")
)
Changing the labels
Finally remove the box around the labels
p_final <- (
ggplot(reefton_data_long,aes(as.Date(Date.local),data)) +
geom_point(aes(colour=measure),shape=1) +
facet_grid(measure~., scales = "free_y", switch="y",
labeller = label_parsed) +
ylab("") +
xlab("Date") +
theme_bw() +
theme(legend.position="none",strip.placement = "outside",
strip.background = element_blank()) + #no background
scale_x_date(date_labels = "%B %Y")
)
Finally remove the box around the labels
Compare back to the initial default graph
